This proposal is a biophysical study of molecular recognition using computer simulation and spectroscopic methods. It aims to (1) predict the thermodynamic properties of peptide binding by glycopeptide antibiotics using computer simulation; (2) make experimental measurements of the relative free energy, enthalpy, entropy, and heat capacity changes upon binding to assess the accuracy of the simulation predictions; and (3) refine the simulations as needed to reconcile discrepancies between simulation and experiment. The eventual goal is to gain insight into the microscopic details governing specific molecular recognition and facilitate the rational design of drugs in this class. These antibiotics are vitally important pharmaceutical agents that have defied chemical modification into clinically useful variants. Their side effect profiles and therapeutic indices are clinically problematic, pointing to the need for alternative and improved substitutes. They are excellent model systems for the study of specific recognition in other and larger systems by virtue of their size, chemical nature, and specificity.